OpenCV  4.2.0
Open Source Computer Vision
Classes | Enumerations | Functions
cv::stereo Namespace Reference

Classes

struct  CensusKernel
 A kernel in which a pixel is compared with the center of the window. More...
 
class  CombinedDescriptor
 
struct  Match
 
class  Matching
 
struct  MCTKernel
 
class  MeanKernelIntegralImage
 
struct  ModifiedCsCensus
 
struct  MVKernel
 
struct  PropagationParameters
 
class  QuasiDenseStereo
 Class containing the methods needed for Quasi Dense Stereo computation. More...
 
class  StarKernelCensus
 implementation for the star kernel descriptor More...
 
class  StereoBinaryBM
 Class for computing stereo correspondence using the block matching algorithm, introduced and contributed to OpenCV by K. Konolige. More...
 
class  StereoBinarySGBM
 The class implements the modified H. Hirschmuller algorithm [102] that differs from the original one as follows: More...
 
class  StereoMatcher
 Filters off small noise blobs (speckles) in the disparity map. More...
 
class  SymetricCensus
 paralel implementation of the center symetric census More...
 

Enumerations

enum  {
  CV_DENSE_CENSUS,
  CV_SPARSE_CENSUS,
  CV_CS_CENSUS,
  CV_MODIFIED_CS_CENSUS,
  CV_MODIFIED_CENSUS_TRANSFORM,
  CV_MEAN_VARIATION,
  CV_STAR_KERNEL
}
 
enum  {
  CV_SPECKLE_REMOVAL_ALGORITHM,
  CV_SPECKLE_REMOVAL_AVG_ALGORITHM
}
 speckle removal algorithms. These algorithms have the purpose of removing small regions More...
 
enum  {
  CV_QUADRATIC_INTERPOLATION,
  CV_SIMETRICV_INTERPOLATION
}
 subpixel interpolationm methods for disparities. More...
 

Functions

void censusTransform (const cv::Mat &image1, const cv::Mat &image2, int kernelSize, cv::Mat &dist1, cv::Mat &dist2, const int type)
 
void censusTransform (const cv::Mat &image1, int kernelSize, cv::Mat &dist1, const int type)
 
void imageMeanKernelSize (const cv::Mat &img, int windowSize, cv::Mat &c)
 
void modifiedCensusTransform (const cv::Mat &img1, const cv::Mat &img2, int kernelSize, cv::Mat &dist1, cv::Mat &dist2, const int type, int t=0, const cv::Mat &IntegralImage1=cv::Mat::zeros(100, 100, CV_8UC1), const cv::Mat &IntegralImage2=cv::Mat::zeros(100, 100, CV_8UC1))
 
void modifiedCensusTransform (const cv::Mat &img1, int kernelSize, cv::Mat &dist, const int type, int t=0, const cv::Mat &IntegralImage=cv::Mat::zeros(100, 100, CV_8UC1))
 
void starCensusTransform (const cv::Mat &img1, const cv::Mat &img2, int kernelSize, cv::Mat &dist1, cv::Mat &dist2)
 
void starCensusTransform (const cv::Mat &img1, int kernelSize, cv::Mat &dist)
 
void symetricCensusTransform (const cv::Mat &img1, const cv::Mat &img2, int kernelSize, cv::Mat &dist1, cv::Mat &dist2, const int type)
 
void symetricCensusTransform (const cv::Mat &img1, int kernelSize, cv::Mat &dist1, const int type)
 

Enumeration Type Documentation

◆ anonymous enum

anonymous enum
Enumerator
CV_DENSE_CENSUS 
CV_SPARSE_CENSUS 
CV_CS_CENSUS 
CV_MODIFIED_CS_CENSUS 
CV_MODIFIED_CENSUS_TRANSFORM 
CV_MEAN_VARIATION 
CV_STAR_KERNEL 

Function Documentation

◆ censusTransform() [1/2]

void cv::stereo::censusTransform ( const cv::Mat image1,
const cv::Mat image2,
int  kernelSize,
cv::Mat dist1,
cv::Mat dist2,
const int  type 
)

Two variations of census applied on input images Implementation of a census transform which is taking into account just the some pixels from the census kernel thus allowing for larger block sizes

◆ censusTransform() [2/2]

void cv::stereo::censusTransform ( const cv::Mat image1,
int  kernelSize,
cv::Mat dist1,
const int  type 
)

◆ imageMeanKernelSize()

void cv::stereo::imageMeanKernelSize ( const cv::Mat img,
int  windowSize,
cv::Mat c 
)

◆ modifiedCensusTransform() [1/2]

void cv::stereo::modifiedCensusTransform ( const cv::Mat img1,
const cv::Mat img2,
int  kernelSize,
cv::Mat dist1,
cv::Mat dist2,
const int  type,
int  t = 0,
const cv::Mat IntegralImage1 = cv::Mat::zeros(100, 100, CV_8UC1),
const cv::Mat IntegralImage2 = cv::Mat::zeros(100, 100, CV_8UC1) 
)

STANDARD_MCT - Modified census which is memorizing for each pixel 2 bits and includes a tolerance to the pixel comparison MCT_MEAN_VARIATION - Implementation of a modified census transform which is also taking into account the variation to the mean of the window not just the center pixel

◆ modifiedCensusTransform() [2/2]

void cv::stereo::modifiedCensusTransform ( const cv::Mat img1,
int  kernelSize,
cv::Mat dist,
const int  type,
int  t = 0,
const cv::Mat IntegralImage = cv::Mat::zeros(100, 100, CV_8UC1) 
)

◆ starCensusTransform() [1/2]

void cv::stereo::starCensusTransform ( const cv::Mat img1,
const cv::Mat img2,
int  kernelSize,
cv::Mat dist1,
cv::Mat dist2 
)

◆ starCensusTransform() [2/2]

void cv::stereo::starCensusTransform ( const cv::Mat img1,
int  kernelSize,
cv::Mat dist 
)

◆ symetricCensusTransform() [1/2]

void cv::stereo::symetricCensusTransform ( const cv::Mat img1,
const cv::Mat img2,
int  kernelSize,
cv::Mat dist1,
cv::Mat dist2,
const int  type 
)

The classical center symetric census A modified version of cs census which is comparing a pixel with its correspondent after the center

◆ symetricCensusTransform() [2/2]

void cv::stereo::symetricCensusTransform ( const cv::Mat img1,
int  kernelSize,
cv::Mat dist1,
const int  type 
)